Unsteady Flow and Transition Phenomena in an Axial Flow ...
Transcript of Unsteady Flow and Transition Phenomena in an Axial Flow ...
Unsteady Flow and Transition
Phenomena in an Axial Flow
Compressor
by
A. D. Henderson, B.E. (Hons)
School of Engineering
Submitted in fulfilment of the requirements
for the degree of
Doctor of Philosophy
University of Tasmania
November 2006
Declaration of Originality
This thesis contains no material which has been accepted for a degree or diploma by the
University or any other institution, except by way of background information and duly
acknowledged in the thesis, and to the best of my knowledge and belief no material
previously published or written by another person except where due acknowledgement
is made in the text of the thesis, nor does the thesis contain any material that infringes
copyright.
Statement of Authority of Access
This thesis is not to be made available for loan or copying for two years following the
date this statement was signed. Following that time the thesis may be made available
for loan and limited copying in accordance with the Copyright Act 1968.
Alan D. Henderson
Abstract
The unsteady mid-span aerodynamics of an outlet stator row in a 1.5-stage low-speed
axial compressor is investigated experimentally and numerically. Two stator blade
rows with characteristically different blade profiles are studied: one of standard British
C4 section and a controlled diffusion (CD) blade with a circular arc leading edge
profile.
A turbulence grid placed at compressor inlet is used to generate turbulence char-
acteristics similar to those occurring in an embedded stage in a multi-stage axial com-
pressor. The stator inlet flow is studied using hot-wire anemometry and compared
with previous measurements made in the natural low inlet turbulence configuration of
the research compressor. Increased turbulence level enhances the dispersion of inlet
guide vane (IGV) wakes. This modifies the interaction between IGV and rotor blade
wakes, leading to a more circumferentially uniform flow field at entry to the stator
with significantly lower periodic unsteadiness.
Laminar-turbulent transition on a C4 stator blade is studied using an array of
surface-mounted hot-film sensors. Comparisons with measurements made at low inlet
turbulence show that the increased inlet turbulence level reduces the extent of peri-
odic transitional flow on the stator blade surface. The blade element behaviour flow
behaviour at high inlet turbulence closely resembles the low inlet turbulence case with
the stator immersed in IGV wake turbulence.
The circular arc leading edge profile of the CD stator produces rapid acceleration
and deceleration at the stator leading edge. The influence of this velocity spike on the
stator boundary layer development and transitional flow behaviour is studied using
an array of surface mounted hot-film sensors. A region of favourable pressure gradient
on the suction surface following the leading edge spike has a stabilising effect on the
boundary layer, with a large region of flow in a laminar or transitional state. Turbulent
spots and instability phenomena in this region are examined for convection speed,
Abstract iv
growth rate and evidence of relaminarisation. In contrast, the flow on the pressure
surface becomes turbulent near the leading edge. The study shows that compressor
blade leading edge profiles have a major influence on boundary layer development over
the whole surface.
The effect of upstream rotor wake passing on the stability of stator blade boundary
layers is examined. The unsteady quasi-three dimensional flow solver, UNSFLO, is
used to interpret surface hot-film data and unsteady laminar flow behaviour at the
leading edge of both C4 and CD stators. Rotor wake chopping is found to stabilise the
pressure surface boundary layer and destabilise the suction surface boundary layer.
Examination of hot-film data points to the leading edge as the principal receptivity
site for transitional flow phenomena occurring on the suction surface of both the C4
and CD blading.
Acknowledgements
I would like to thank my colleagues, friends and family for their support throughout
my PhD candidature. In particular, I would like to acknowledge help from a number
of people.
The project would not have been possible without the encouragement and financial
support from the project sponsor, Rolls-Royce Plc. The helpful advice and comments
of Jeremy Hughes, Mike Howard and John Coupland were very much appreciated.
The staff at the School of Engineering Workshop assisted in preparing the research
compressor and instrumentation for experimental measurements. Peter Seward pa-
tiently spent many weeks meticulously instrumenting compressor blades. Glen May-
hew persevered in isolating electrical noise from the research compressor. John Mc-
Culloch assisted in setting up and maintaining the data acquisition systems. Ray
LeFevre and Nathan Smith provided assistance on numerous occasions.
Many thoughtful discussions with other colleges at the School including Jane Sar-
gison and Michael Kirkpatrick were very much appreciated.
Much gratitude is extended to my supervisor, Greg Walker who has given me
helpful advice throughout the project. His thoughtfulness and understanding have
been nothing short of inspirational. I could not have wished for a better supervisor.
In the early stages of the project, Frank Sainsbury and Jason Roberts encouraged
me to learn and use Unix and Linux operating systems and document preparation
in LATEX. These tools made many aspects of my project and preparation of this
manuscript much easier than would have otherwise been.
I would like to thank Ruth and James for their continuous love and support. Ruth
has been very patient and understanding of the long hours devoted to this thesis. Such
kindness I will never forget.
Finally, I would like to thank my parents and other family members for their
unending love and support.
Alan D. Henderson
Contents
Abstract iii
Acknowledgements v
List of Figures xi
List of Tables xx
Nomenclature xxii
1 Introduction 1
2 Literature Review 6
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 The Modes of Transition . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.1 Natural Transition . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.2 Bypass Transition . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.3 Separated Flow Transition . . . . . . . . . . . . . . . . . . . . . 10
2.2.4 Relaminarisation . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3 Predicting the Onset of Transition . . . . . . . . . . . . . . . . . . . . 14
2.4 Linear Stability Theory . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.5 Turbulent Spot Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.6 Transitional Flow in Turbomachinery . . . . . . . . . . . . . . . . . . . 23
2.6.1 Periodic Unsteady Transition . . . . . . . . . . . . . . . . . . . 24
2.6.2 Wake Jet Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.6.3 Separation Bubbles . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Contents vii
3 Experimental Facilities 30
3.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.2 Research Compressor Description . . . . . . . . . . . . . . . . . . . . . 32
3.3 Compressor Detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.3.1 C4 Compressor Blading . . . . . . . . . . . . . . . . . . . . . . 35
3.3.2 CD Stator Blade Row . . . . . . . . . . . . . . . . . . . . . . . 36
3.4 Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.4.1 Pressure Measurement . . . . . . . . . . . . . . . . . . . . . . . 40
3.4.2 Relative Humidity Probe . . . . . . . . . . . . . . . . . . . . . 40
3.4.3 Thermometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.4.4 Multimeter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.4.5 Constant Temperature Anemometers . . . . . . . . . . . . . . . 41
3.4.6 Hot-Wire Probes . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.4.7 High-Speed Data Acquisition . . . . . . . . . . . . . . . . . . . 43
3.4.8 Three-Hole Probe . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.5 Research Compressor Control . . . . . . . . . . . . . . . . . . . . . . . 47
3.5.1 Speed Control . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.5.2 Flow Coefficient . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.6 Turbulence Generating Grid . . . . . . . . . . . . . . . . . . . . . . . . 49
4 Experimental Approach 50
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.2 Inlet Turbulence Level . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.3 Blade Row Clocking . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.4 Compressor Loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.5 Reynolds Number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.6 Total Pressure and Flow Angle . . . . . . . . . . . . . . . . . . . . . . 57
4.7 Blade Surface Pressure Measurements . . . . . . . . . . . . . . . . . . 60
4.8 Stator Inlet Reynolds Number . . . . . . . . . . . . . . . . . . . . . . . 61
4.9 Presentation of Hot-Wire Measurements . . . . . . . . . . . . . . . . . 61
4.10 Presentation of Hot-Film Measurements . . . . . . . . . . . . . . . . . 63
4.11 Blade Surface Particle Trajectories . . . . . . . . . . . . . . . . . . . . 65
5 Stator Inlet Flow 66
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Contents viii
5.2 Experimental Technique . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.3 Unsteady Flow Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.3.1 Turbulence Grid–IGV Blade Row Clocking Effects . . . . . . . 75
5.3.2 Ensemble Averaged Velocity-Time Traces . . . . . . . . . . . . 75
5.4 Integral Length Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
6 C4 Stator Observations 81
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.2 Range of Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
6.3 Time-Mean Stator Inlet Flow . . . . . . . . . . . . . . . . . . . . . . . 85
6.4 Surface Velocity Distributions . . . . . . . . . . . . . . . . . . . . . . . 85
6.5 Hot-Film Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.5.1 Hot-Film Sensor Array . . . . . . . . . . . . . . . . . . . . . . . 87
6.5.2 Measurement Technique . . . . . . . . . . . . . . . . . . . . . . 88
6.5.3 Surface Intermittency Distributions . . . . . . . . . . . . . . . 89
6.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
7 CD Stator Observations 99
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
7.2 Circular Arc Leading Edge Profiles . . . . . . . . . . . . . . . . . . . . 100
7.3 Preliminary Study of the CD Stator . . . . . . . . . . . . . . . . . . . 104
7.3.1 Modelling Approach . . . . . . . . . . . . . . . . . . . . . . . . 105
7.3.2 Predicted Surface Pressure Distributions . . . . . . . . . . . . . 105
7.4 Range of Experimental Investigation . . . . . . . . . . . . . . . . . . . 109
7.5 Surface Pressure Distribution . . . . . . . . . . . . . . . . . . . . . . . 111
7.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
7.5.2 Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . 111
7.5.3 Experimental Method . . . . . . . . . . . . . . . . . . . . . . . 112
7.5.4 Surface Velocity Predictions . . . . . . . . . . . . . . . . . . . . 113
7.5.5 Surface Velocity Distributions . . . . . . . . . . . . . . . . . . . 114
7.6 Hot-Film Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
7.6.1 Hot-Film Sensor Array . . . . . . . . . . . . . . . . . . . . . . . 119
7.6.2 Measurement Technique . . . . . . . . . . . . . . . . . . . . . . 120
7.6.3 Surface Intermittency Distributions . . . . . . . . . . . . . . . 121
Contents ix
7.6.4 General Observations . . . . . . . . . . . . . . . . . . . . . . . 126
7.6.5 Incidence Effects . . . . . . . . . . . . . . . . . . . . . . . . . . 136
7.6.6 Reynolds Number Effects . . . . . . . . . . . . . . . . . . . . . 137
7.7 Stator Wake Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
7.7.1 Experimental Method . . . . . . . . . . . . . . . . . . . . . . . 138
7.7.2 Stator Outlet Flow Field . . . . . . . . . . . . . . . . . . . . . 139
7.7.3 Time-Mean Stator Pressure Loss . . . . . . . . . . . . . . . . . 142
7.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
8 Unsteady Transition Phenomena at the Leading Edge of Compressor
Blades 148
8.1 Scope of Experimental Investigation . . . . . . . . . . . . . . . . . . . 150
8.2 Unsteady Flow Simulation . . . . . . . . . . . . . . . . . . . . . . . . . 152
8.2.1 The UNSFLO Solver . . . . . . . . . . . . . . . . . . . . . . . . 152
8.2.2 Model Description . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.2.3 Model Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
8.2.4 Predicted Leading Edge Boundary Layer Behaviour . . . . . . 159
8.2.5 Effect of Varying Rotor–Stator Axial Spacing . . . . . . . . . . 167
8.3 Transitional Flow Observations . . . . . . . . . . . . . . . . . . . . . . 173
8.3.1 C4 Stator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
8.3.2 CD Stator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
8.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
8.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
9 Conclusions and Recommendations for Future Research 182
A Stator Blade Instrumentation 187
A.1 C4 Stator Blade Instrumentation . . . . . . . . . . . . . . . . . . . . . 187
A.2 CD Stator Blade Instrumentation . . . . . . . . . . . . . . . . . . . . . 191
B Design of the Turbulence Grid 195
B.1 Turbulence Intensity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
B.2 Integral Length Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
B.3 Pressure Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Contents x
C Compressor Reference Pressure 198
C.1 Model of Research Compressor Inlet Contraction . . . . . . . . . . . . 198
C.2 Calibration of Inlet Contraction . . . . . . . . . . . . . . . . . . . . . . 202
Bibliography 203
List of Figures
2.1 Smoke visualisation of vortex trusses (from Knapp and Roache [96]) . 8
2.2 Schematic diagram of the natural transition process (from White [188]) 8
2.3 Visualisation of DNS solution showing streaks proceeding turbulent
breakdown (from Jacobs and Durbin [90]) . . . . . . . . . . . . . . . . 10
2.4 Schematic diagram of separated flow transition (from Mayle [111]) . . 11
2.5 Disturbance amplification curves for the boundary layer on a flat plate
at zero incidence. δ1 denotes boundary layer displacement thickness
(from Schlichting [147]). . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.6 Comparison of critical Reθ for turbulent breakdown for both the en and
Abu Ghannam and Shaw [1] methods (from Drela [35]). Critical ampli-
fication ratio, turbulence intensity, and momentum thickness Reynolds
number are indicated by Ncrit, t and Rtheta, respectively. . . . . . . 19
2.7 Schematic diagram of turbulent spot development (from Paxson and
Mayle [129]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.8 The domain of dependence for a turbulent spot (from Mayle [111]) . . 21
2.9 Schematic showing the morphology of periodic unsteady transition on
a compressor suction surface (from Halstead et al. [61]) . . . . . . . . 24
2.10 Interaction between a convecting wake and flat plate (adapted from
Hodson [73] after Meyer [115]) . . . . . . . . . . . . . . . . . . . . . . 26
2.11 Vectors and streamlines showing the interaction between a negative jet
and a suction side boundary layer on a LP turbine blade (from Hodson
and Howell [75]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.12 Schematic showing leading edge separation bubble (from Tain and Cump-
sty [167]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
List of Figures xii
3.1 UTAS research compressor with the compressor section open showing
the rotor and stator blade rows . . . . . . . . . . . . . . . . . . . . . . 31
3.2 Schematic diagram of the research compressor (adapted from Oliver
[122]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.3 The research compressor showing original C4 rotor and stator blade
rows (left) and the CD stator blade row with rotor disk removed (right) 35
3.4 Radial view of C4 and CD stator blades showing hub, mid-span and
tip sections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.5 United Sensor CA120 three-hole probe . . . . . . . . . . . . . . . . . . 44
3.6 United Sensor CA120 three-hole probe calibration . . . . . . . . . . . 46
3.7 Photograph of the research compressor inlet section showing the tur-
bulence grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.1 Cross-section of the research compressor showing mid-span blade row
configuration and typical instantaneous wake dispersion pattern (C4
stator) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.2 Reynolds number variation in the PW2037 high bypass ratio turbofan
at altitude cruise (from Hourmouziadis [79]) . . . . . . . . . . . . . . . 56
4.3 Pitchwise variation of total pressure coefficient in the rotor–stator ax-
ial gap at mid-span. Three-hole probe measurements 20.7%c axially
upstream from the stator leading edge (from Hughes [83], Rec = 120000) 58
4.4 Pitchwise variation of total pressure coefficient with grid–IGV align-
ment in the rotor–stator axial gap at mid-span height. Total pressure
tube measurements at 18.7%c axially upstream from the CD stator
leading edge (based on C4 chord length, Rec = 120000, φ = 0.675) . . 59
5.1 View looking upstream showing the path of the probe relative to the
IGV blade row . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.2 Unsteady flow field in the rotor–stator axial space at low compressor
load (φ = 0.840, and Rec = 120000). Top: low turbulence case with-
out grid. Bottom: high turbulence case with grid (grid wakes in IGV
passage g/S = 0.5). Filled contours show ensemble-averaged velocity
〈u〉/us. Line contours show ensemble-averaged turbulence level 〈Tu〉 in
1% intervals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
List of Figures xiii
5.3 Unsteady flow field in the rotor–stator axial space at medium compres-
sor load (φ = 0.675, and Rec = 120000). Top: low turbulence case
without grid. Bottom: high turbulence case with grid (grid wakes in
IGV passage g/S = 0.5). Filled contours show ensemble-averaged ve-
locity 〈u〉/us. Line contours show ensemble-averaged turbulence level
〈Tu〉 > in 1% intervals. . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.4 Unsteady flow field in the rotor-stator axial space at high compressor
load (φ = 0.600 and Rec = 120000). Top: low turbulence case with-
out grid. Bottom: high turbulence case with grid (grid wakes in IGV
passage g/S = 0.5). Filled contours show ensemble-averaged velocity
〈u〉/us. Line contours show ensemble-averaged turbulence level 〈Tu〉 >in 1% intervals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.5 Unsteady flow field in the rotor–stator axial space at medium compres-
sor load with the turbulence grid (φ = 0.675 and Rec = 120000). Top:
grid wakes incident on IGV blade row (g/S = 0.0). Bottom: grid wakes
in IGV passage (g/S = 0.5). Filled contours show ensemble-averaged
velocity 〈u〉/us. Line contours show ensemble-averaged turbulence level
〈Tu〉 > in 1% intervals. . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.6 Temporal variation in velocity at several circumferential positions for
medium load (φ = 0.675 and Rec = 120000). Common time origin
t∗ = 0.5 at centre of rotor wake disturbance. . . . . . . . . . . . . . . . 77
6.1 C4 stator surface velocity distribution at mid-span (Rec = 120000) for
LAG configuration. Experimental data from Solomon [154] . . . . . . 84
6.2 C4 stator suction surface velocity distributions at mid-span (Rec =
120000). Measurements without grid from Hughes [83]. . . . . . . . . . 86
6.3 The C4 stator instrumented with an array of surface hot-film sensors . 88
6.4 C4 stator surface intermittency distribution at high loading (φ = 0.600,
Rec = 120000). Colour filled contours show ensemble average inter-
mittency (〈γ〉), line contours show probability of relaxing flow in 1%
intervals, and particle trajectories at 1.0U, 0.88U, 0.7U and 0.5U . . . 90
List of Figures xiv
6.5 C4 stator surface intermittency distribution at medium loading (φ =
0.675, Rec = 120000). Colour filled contours show ensemble average
intermittency (〈γ〉), line contours show probability of relaxing flow in
1% intervals, and particle trajectories at 1.0U, 0.88U, 0.7U and 0.5U . 91
6.6 C4 stator surface intermittency distribution at low loading (φ = 0.840,
Rec = 120000). Colour filled contours show ensemble average inter-
mittency (〈γ〉), line contours show probability of relaxing flow in 1%
intervals, and particle trajectories at 1.0U, 0.88U, 0.7U and 0.5U . . . 92
6.7 C4 stator surface intermittency distribution at medium loading with
IGV wakes in stator blade row passage (a/S = 0.5, φ = 0.675 and Rec =
120000). Colour filled contours show ensemble average intermittency
(〈γ〉), line contours show probability of relaxing flow in 1% intervals,
particle trajectories at 1.0U, 0.88U, 0.7U and 0.5U . . . . . . . . . . . 96
6.8 Time-mean ensemble averaged intermittency distribution for the C4
stator (γ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7.1 Predicted CD stator surface pressure distributions at various incidences
(MISES, Tu=4%, Re1=230000) . . . . . . . . . . . . . . . . . . . . . . 106
7.2 Predicted exit angle (top), pressure loss coefficient (centre), and di-
mensionless surface length to transition onset on the suction surface
(bottom) from the MISES flow solver (Tu=4.0% and Re1=230000) . . 108
7.3 CADKEY model of the CD stator blade with mounting boss showing
milling paths for pressure tube tracks (suction surface view) . . . . . . 112
7.4 CD stator surface velocity distributions. Experimental results and
MISES predictions for incidence Cases A and B . . . . . . . . . . . . . 115
7.5 CD stator surface velocity distributions. Experimental results and
MISES predictions for incidence Cases C and D. . . . . . . . . . . . . 116
7.6 CD stator surface velocity distributions around leading edge. Experi-
mental results and MISES predictions for Cases A–D . . . . . . . . . . 117
7.7 CD stator blade with array of surface mounted hot-film sensors installed
in the UTAS research compressor . . . . . . . . . . . . . . . . . . . . . 119
List of Figures xv
7.8 CD stator surface intermittency distributions for incidence Case A.
Colour contours show ensemble average intermittency (〈γ〉); line con-
tours show probability of calmed flow in intervals of 0.1; white lines
show particle trajectories at 1.0U, 0.88U, 0.7U and 0.5U . . . . . . . . 122
7.9 CD stator surface intermittency distributions for incidence Case B.
Colour contours show ensemble average intermittency (〈γ〉); line con-
tours show probability of calmed flow in intervals of 0.1; white lines
show particle trajectories at 1.0U, 0.88U, 0.7U and 0.5U . . . . . . . . 123
7.10 CD stator surface intermittency distributions for incidence Case C.
Colour contours show ensemble average intermittency (〈γ〉); line con-
tours show probability of calmed flow in intervals of 0.1; white lines
show particle trajectories at 1.0U, 0.88U, 0.7U and 0.5U . . . . . . . . 124
7.11 CD stator surface intermittency distributions for incidence Case D.
Colour contours show ensemble average intermittency (〈γ〉); line con-
tours show probability of calmed flow in intervals of 0.1; white lines
show particle trajectories at 1.0U, 0.88U, 0.7U and 0.5U . . . . . . . . 125
7.12 Temporal variation of ensemble average turbulent intermittency (〈γ〉)(colour contours) around the CD stator leading edge (Rec = 120000) . 126
7.13 Hot-film measurements near the CD stator trailing edge (Case D, Rec =
120000). Top: colour contours of ensemble average intermittency (〈γ〉)with line contours of probability of calmed flow in intervals of 0.1. Bot-
tom: typical set of simultaneously acquired quasi wall shear stress mea-
surements. Sensor locations are indicated by 5. Lines show particle
trajectories at speeds of 1.0U , 0.88, 0.7U and 0.5U . . . . . . . . . . . . 129
7.14 CD stator hot-film traces at Rec = 95000 . . . . . . . . . . . . . . . . 130
7.15 CD stator hot-film traces at Rec = 120000 . . . . . . . . . . . . . . . . 131
7.16 CD stator hot-film traces at Rec = 160000 . . . . . . . . . . . . . . . . 132
7.17 Predicted acceleration parameter and momentum thickness Reynolds
number for the CD stator (MISES, Cases A and B) . . . . . . . . . . 134
7.18 Predicted acceleration parameter and momentum thickness Reynolds
number for the CD stator (MISES, Cases C and D) . . . . . . . . . . . 135
List of Figures xvi
7.19 Unsteady flow field 19.7%c axially downstream from CD stator trailing
edge. Top: Case A (φ = 0.750). Bottom: Case B (φ = 0.710). Left:
time-mean ensemble averaged velocity. Centre: time-mean random,
periodic and total disturbance level. Right: colour contours of ensemble
averaged velocity; line contours of ensemble average turbulence Tu with
1% interval spacing (Rec = 120000) . . . . . . . . . . . . . . . . . . . . 140
7.20 Unsteady flow field 19.7%c axially downstream from CD stator trailing
edge. Top: Case C (φ = 0.675). Bottom: Case D (φ = 0.600). Left:
time-mean ensemble averaged velocity. Centre: time-mean random,
periodic and total disturbance level. Right: colour contours of ensemble
averaged velocity; line contours of ensemble average turbulence Tu with
1% interval spacing (Rec = 120000) . . . . . . . . . . . . . . . . . . . . 141
7.21 Time-mean wake velocity profile of a CD stator blade showing estimated
linear inviscid velocity distribution (Case A, φ = 0.750, Rec = 120000) 143
7.22 Variation of pitchwise average stator exit flow angle (α2) with incidence
(top); variation of total pressure loss coefficient (ω) with incidence (cen-
tre); and variation in predicted suction surface transition onset location
(MISES) and extent of suction surface transitional flow (0.2 < γ < 0.8)
with incidence (bottom). All measurements made at Rec = 120000. . . 145
8.1 Unsteady flow field around C4 stator leading edge (UNSFLO) for LAG
configuration (Case A, φ = 0.675, Rec = 120000). Vectors indicate
perturbation from local time-mean velocity. Colour shading indicates
fluid entropy relative to the maximum level at the wake centre . . . . 156
8.2 Unsteady flow field around C4 stator leading edge (UNSFLO) for SAG
configuration (Case C, φ = 0.675, Rec = 120000). Vectors indicate
perturbation from local time-mean velocity. Colour shading indicates
fluid entropy relative to the maximum level at the wake centre . . . . 157
8.3 Unsteady flow field around CD stator leading edge (UNSFLO) for SAG
configuration (Case E, φ = 0.675, Rec = 120000). Vectors indicate
perturbation from local time-mean velocity. Colour shading indicates
fluid entropy relative to the maximum level at the wake centre . . . . 158
List of Figures xvii
8.4 Comparison of dimensionless ensemble-averaged quasi wall shear stress
〈τq〉 /τq from hot-film sensors (left) and dimensionless skin friction fac-
tor cf/cf from UNSFLO (right). C4 stator, LAG configuration, Case
A, φ = 0.675, Rec = 120000 . . . . . . . . . . . . . . . . . . . . . . . . 160
8.5 Comparison of dimensionless ensemble-averaged quasi wall shear stress
〈τq〉 /τq from hot-film sensors (left) and dimensionless skin friction fac-
tor cf/cf from UNSFLO (right). C4 stator, SAG configuration, Case
C, φ = 0.675, Rec = 120000 . . . . . . . . . . . . . . . . . . . . . . . . 161
8.6 Comparison of dimensionless ensemble-averaged quasi wall shear stress
〈τq〉 /τq from hot-film sensors (left) and dimensionless skin friction fac-
tor cf/cf from UNSFLO (right). CD stator, SAG configuration, Case
E, φ = 0.675, Rec = 120000 . . . . . . . . . . . . . . . . . . . . . . . . 162
8.7 Temporal variation in momentum thickness Reynolds number and shape
factor from UNSFLO computations. C4 stator, LAG configuration
(Case A, φ = 0.675, Rec = 120000) . . . . . . . . . . . . . . . . . . . . 163
8.8 Temporal variation in momentum thickness Reynolds number and shape
factor from UNSFLO computations. C4 stator, SAG configuration
(Case C, φ = 0.675, Rec = 120000) . . . . . . . . . . . . . . . . . . . . 164
8.9 Temporal variation in momentum thickness Reynolds number and shape
factor from UNSFLO computations. CD stator, SAG configuration
(Case E, φ = 0.675, Rec = 120000) . . . . . . . . . . . . . . . . . . . . 165
8.10 C4 stator surface ensemble-averaged intermittency (colour contours)
and probability of instability wave occurrence (line contours) with su-
perimposed particle trajectories at different proportions of local free-
stream velocity (adapted from Hughes and Walker [85]) and added neu-
tral stability boundary predicted by the present study (Case A, LAG,
Rec = 120000) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
8.11 C4 stator surface ensemble-averaged intermittency (colour contours)
and probability of instability wave occurrence (line contours) with su-
perimposed particle trajectories at different proportions of local free-
stream velocity (adapted from Hughes and Walker [85]) and added neu-
tral stability boundary predicted by the present study (Case B, LAG,
Rec = 120000) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
List of Figures xviii
8.12 C4 stator surface ensemble-averaged intermittency (colour contours)
and probability of instability wave occurrence (line contours) with su-
perimposed particle trajectories at different proportions of local free-
stream velocity and added neutral stability boundary predicted by the
present study (Case C, SAG, Rec = 120000) . . . . . . . . . . . . . . . 170
8.13 C4 stator surface ensemble-averaged intermittency (colour contours)
and probability of instability wave occurrence (line contours) with su-
perimposed particle trajectories at different proportions of local free-
stream velocity and added neutral stability boundary predicted by the
present study (Case D, SAG, Rec = 120000) . . . . . . . . . . . . . . . 171
8.14 CD stator surface ensemble-averaged intermittency (colour contours)
and probability of instability wave occurrence (line contours) with su-
perimposed particle trajectories at different proportions of local free-
stream velocity and neutral stability boundary predicted by the present
study (Case E, SAG, Rec = 120000) . . . . . . . . . . . . . . . . . . . 172
8.15 Typical raw quasi wall shear stress traces on C4 stator surface (LAG
configuration). Top: IGV wakes aligned in stator passage (Case A).
Bottom: IGV wakes incident on stator blade row (Case B). LAG, Rec =
120000. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
8.16 Typical raw quasi wall shear stress traces on C4 stator surface (SAG
configuration). Top: IGV wakes aligned in stator passage (Case C).
Bottom: IGV wakes incident on stator blade row (Case D). SAG, Rec =
120000. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
8.17 Typical raw quasi wall shear stress traces on CD stator surface (Case
E, SAG, Rec = 120000) . . . . . . . . . . . . . . . . . . . . . . . . . . 178
A.1 Mid-span pressure tapping locations of C4 stator blade (top) and mid-
span hot-film sensor locations of C4 stator blade (bottom) . . . . . . . 188
A.2 Mid-span surface pressure tapping locations of CD stator blade (top)
and mid-span hot-film sensor locations of CD stator blade (bottom) . 192
C.1 Numerical simulation of research compressor inlet contraction (CFX).
Rendered view of intake model (top) with corresponding contours of
pressure coefficient on a radial plane and pressure coefficient of intake
surfaces with and without loss terms for the grid included (bottom) . 201
List of Figures xix
C.2 Calibration of research compressor contraction pressure coefficient with-
out turbulence grid installed (markers). Lines indicate flow coefficient
calculated by the method presented above . . . . . . . . . . . . . . . . 203
List of Tables
3.1 Compressor blade details for studies of Solomon [154], Hughes [83], and
present investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.1 Pitchwise averaged integral length scale for low and high turbulence
cases (medium loading φ = 0.675) . . . . . . . . . . . . . . . . . . . . . 79
6.1 Circumferentially averaged C4 stator inlet incidence with corresponding
stator inlet Reynolds number at low and high inlet turbulence . . . . . 85
7.1 Matrix of measured operating conditions for the CD stator blade row:
pitchwise average incidence (i◦ ) and inlet Reynolds number (Re1) at
entry to the CD stator blade row for various values of flow coefficient
(φ) and reference Reynolds number (Rec) . . . . . . . . . . . . . . . . 110
8.1 Matrix of test conditions for the study of the flow around the leading
edge of the C4 and CD stator blades. All measurements were made at
constant reference Reynolds number (Rec = 120000) . . . . . . . . . . 151
A.1 Mid-span pressure tapping locations of C4 stator blade suction sur-
face. All coordinates are relative to the geometrical blade leading
edge defined as the intersection of the leading edge and camber line
(x = y = x∗ = s∗ = 0). x∗ = x/cx is dimensionless axial distance.
cx = c cos(ξ) is the axial projection of chord length. s∗ = s/smax is
dimensionless surface length. c = 76.2 mm and smax = 79.23 mm
(adapted from Solomon [154]) . . . . . . . . . . . . . . . . . . . . . . . 189
List of Tables xxi
A.2 Mid-span hot-film sensor locations of C4 stator blade. All coordinates
are relative to the geometrical blade leading edge defined as the inter-
section of the leading edge and camber line (x = y = x∗ = s∗ = 0).
x∗ = x/cx is dimensionless axial distance. cx = c cos(ξ) is the ax-
ial projection of chord length. s∗ = s/smax is dimensionless surface
length. c = 76.2 mm, smax = 79.23 mm on the suction surface and
smax = 76.27 on the pressure surface (adapted from Solomon [154]) . . 190
A.3 Mid-span pressure tapping locations of CD stator blade. All coordi-
nates are relative to the geometrical blade leading edge defined as the
intersection of the leading edge and camber line (x = y = x∗ = s∗ = 0).
x∗ = x/cx is dimensionless axial distance. cx = c cos(ξ) is the ax-
ial projection of chord length. s∗ = s/smax is dimensionless surface
length. c = 152.4 mm and smax = 162.1 mm on the suction surface and
smax = 154.8 mm on the pressure surface . . . . . . . . . . . . . . . . 193
A.4 Mid-span hot-film sensor locations of CD stator blade. All coordinates
are relative to the geometrical blade leading edge defined as the inter-
section of the leading edge and camber line (x = y = x∗ = s∗ = 0).
x∗ = x/cx is dimensionless axial distance. cx = c cos(ξ) is the ax-
ial projection of chord length. s∗ = s/smax is dimensionless surface
length. c = 152.4 mm and smax = 162.1 mm on the suction surface and
smax = 154.8 mm on the pressure surface . . . . . . . . . . . . . . . . 194
C.1 Table of CFX model parameters . . . . . . . . . . . . . . . . . . . . . 200
Nomenclature
General Variables
c blade chord
cf skin friction coefficient = 2τw/ρU2
Cp surface pressure coefficient = (P1 − p)/(P1 − p1)
D diameter
E anemometer bridge voltage
Eo anemometer bridge voltage at zero flow = ((E2 −E2o )/E2
o )3
H boundary layer or wake shape factor = δ∗/θ
i blade incidence angle = α− β
K acceleration parameter = (ν/U 2)dU/dx
L length
p static pressure
P total pressure
R radius
Rea compressor inlet Reynolds number = Va.c/ν
Rec compressor reference Reynolds number = Umb.c/ν
Re1 stator inlet Reynolds number = V1.c/ν
s surface length
s∗ dimensionless surface length = s/smax
smax surface length from leading to trailing edge
S blade pitch
S∗ dimensionless pitchwise distance = w/S
t time
t∗ dimensionless time = t/T
T rotor passing period
Tu random disturbance level or ‘turbulence’
Nomenclature xxiii
T̃ u periodic disturbance level
TuD apparent or ‘total’ disturbance level
U local free-stream velocity
Umb mid-blade rotor blade speed
Va axial velocity at compressor inlet
w periperal distance
x axial coordinate
α flow angle relative to axial direction, turbulent spot spreading angle
β blade angle relative to axial axial direction
γ turbulent intermittency
λθ pressure gradient parameter = (θ2/ν)dU/dx
Λ integral or ‘macro’ turbulence scale
ρ air density
φ flow coefficient = Va/Umb
σ blade row solidity = c/S
ν kinematic viscosity = µ/ρ
µ absolute viscosity
τw wall shear stress
τq quasi wall shear stress
θ boundary layer momentum thickness, wake momentum thickness, blade
camber angle = β1 − β2
δ∗ boundary layer displacement thickness, wake displacement thickness
ξ blade stagger angle
ω total pressure loss coefficient = (P1 − P2)/(P1 − p1)
Subscripts
1 stator blade row inlet
2 stator blade row outlet
cr critical
in compressor inlet (upstream from IGV)
le leading edge
mb mid-blade
te trailing edge
Nomenclature xxiv
Acronyms
AVDR axial velocity – density ratio
CD controlled diffusion
CFD computational fluid dynamics
DNS direct numerical simulation
HP high-pressure
LAG long axial gap
LDA laser-doppler anemometry
LDV laser-doppler velocimetry
LES large eddy simulation
LP low-pressure
OD outer diameter
PDF probability density function
PVC peak-valley counting
RANS Reynolds-averaged Navier–Stokes
SAG short axial gap
UTAS University of Tasmania
Mathematical Notation
( ) time average
〈 〉 phase-lock or ensemble averaging